Insect Control: Biological and Synthetic Agents - Index of

364 10: Genetically Modified Baculoviruses for Pest Insect Control
use of AcMLF9.ScathL in comparison to the use of
AcMNPV as a biological pesticide. Smith et al.
(2000a) have investigated the density and diversity
of nontarget predators under field conditions following
the application of recombinant baculoviruses
(AcMNPV or HzSNPV expressing scorpion
toxin LqhIT2 under the ie1 gene promoter) on
cotton. They found that predator densities and
diversity were similar between recombinant and
wild-type baculovirus treated plots. In contrast, the
chemical pesticide (esfenvalerate) treated plots had
consistently lower predator populations.
Taken together, these studies indicate that: (1) the
amount of baculovirus expressed toxin or protease
that accumulates in the larvae is not sufficient to
induce any adverse effects on the predator; (2) the
baculovirus is not infectious towards the predatory
insects; and (3) the toxin or protease encoding gene
is not expressed in the predatory insect. In some
cases, there appear to be some costs associated
with the beneficial insects that prey upon virus
infected larvae. However, these costs are significantly
lower in comparison to the costs associated with
the application of synthetic chemical pesticides. The
development of selective recombinant insecticides
should augment any IPM program by reducing the
impact on nontarget species, including beneficial
insects. Consequently, the resurgence of primary
pests and outbreaks of secondary pests should be
minimized.
10.5.2. Fitness of GM Baculoviruses
Fitness is a term that describes the ability of an
organism to produce progeny that survive to contribute
to the following generation (Cory, 2000). In
order to estimate the relative fitness of a recombinant
baculovirus in comparison to the wild-type
baculovirus, five key parameters should be assessed:
speed of kill, yield, transmission, dispersal, and persistence.
Speed of kill (or time to death), virus yield,
and transmission rate can be easily determined by
laboratory bioassays (as discussed above) and in
some cases under field conditions. Although the
relationship between speed of kill (generally quantified
in terms of LT 50) and virus yield is complex,
faster speed of kill generally results in dramatically
lower virus yields. This correlation between
improved speed of kill and reduced virus yield is
found regardless of the parental virus that is genetically
modified, and for modifications in which an
effector gene is inserted or when an endogenous
gene is deleted from the genome. For example, the
speeds of kill of third, fourth, or fifth instar larvae of
T. ni that are infected with AcAaIT or AcJHE.KK
are reduced by approximately 30% and 8%, respectively,
in comparison to control larvae infected
with AcMNPV. These faster speeds of kill result in
reductions of approximately 80% and 40% in the
yields (polyhedra per milligram of cadaver) of
AcAaIT and AcJHE.KK, respectively, in comparison
to the yield of AcMNPV (Kunimi et al., 1996).
Dramatic reductions of up to 95% in virus yield
(polyhedra per microgram of cadaver) are found in
second and fourth instar larvae of AcTOX34.4
infected T. ni in comparison to control AcMNPV
infected larvae (Burden et al., 2000). The corresponding
reductions in the mean times to death are
50–60%. A reduction in virus yield is also found by
deletion of the egt gene of AgMNPV (Pinedo et al.,
2003). The mean lethal times of third instar larvae
of A. gemmatalis infected with various doses of
vAgEGTD-lacZ is reduced by 10–26% in comparison
to control larvae infected with the same dose.
The yield (polyhedra per gram of cadaver) of
vAgEGTD-lacZ is reduced by approximately 50%
in comparison to control larvae infected with the
wild-type AgMNPV. Similar results are found in
fifth instar larvae of S. frugiperda infected with
vEGTDEL, which produce 23% fewer polyhedra
per insect (the yield of virus per milligram of cadaver
is not, however, significantly different) in comparison
to control larvae infected with AcMNPV
(O’Reilly and Miller, 1991). O’Reilly et al. (1991)
and Ignoffo et al. (2000) speculated that this correlation
between improved speed of kill and reduced
virus yield results from the considerably reduced
size of recombinant baculovirus infected larvae at
the time of death.
Milks et al. (2001) have focused on intrahost
competition between AcAaIT and AcMNPV or
AcAaIT and TnSNPV in larvae of T. ni that were
synchronously or asynchronously infected. They
found no differences in the fitness of the genetically
modified or wild-type viruses in terms of virus yield.
The most important factors in these mixed infections
were dose and timing. The virus that was
inoculated at the highest dose or the virus that
was first inoculated was the one that had the competitive
advantage. These findings are not unreasonable
when one considers that there are no significant
differences in the replication rates of toxin gene
carrying and wild-type baculoviruses in cell culture.
A key component of the transmission rate of a
virus is its pathogenicity or potency (often quantified
in terms of the LD50 or LC50). In general, the
pathogenicity of a baculovirus is not significantly
changed following the insertion of a neurotoxin
gene into its genome (McCutchen et al., 1991;
Tomalski and Miller, 1992; Prikhod’ko et al.,